Theresa Putkey Welcome to this talk on how to use content modeling to define your taxonomy framework. Before we dive in, I'll tell you a little bit about myself. My name is Theresa Putkey and I am a structural content strategist. So I focus on information architecture, taxonomy and metadata, and the strategy around those things. I live in Vancouver, Canada, and I've been in IA and UX since 2004. What we'll cover today is how a text content model reveals metadata and taxonomy needs. Why using a content model to identify taxonomy ensures that you develop the appropriate taxonomy. And I'm going to give you a clear example to anchor the discussion. In the past, I have often fumbled around trying to figure out what taxonomies are useful, the priorities of the taxonomies and how to develop them and I want to help you get past this and figure out what taxonomy is needed based on the content model. Theresa Putkey Before we dive into doing that content modeling work, I want to give you some definitions. In the past, I have done this talk or something similar. And in the end, somebody asked me, I really enjoyed your presentation. But what does taxonomy mean? So instead of going through this, like we all know what we're talking about, I'm going to give you a few definitions so we all have the same understanding. If you know this stuff, this will take just a few minutes. So Information Architecture encompasses sitemaps, content types, wireframes, search design, and it helps standardize information and content on a site. And here I'm using IA in the sense of the deliverable, not in a sense of the profession. So a lot of different people can do information architecture work and create IA deliverables. Here's an example of a sitemap. You can see it has multiple levels, multiple pages, and the relationships, the hierarchical relationships are displayed here. Here's another site map that I did in Excel. The first one I did in Slickplan. It's just a way to lay out the hierarchical structure of pages on the site. Theresa Putkey When I talk about metadata, I simply mean data about data. In a content management system metadata refers to the fields that are on the content pages or on the assets. It includes fields like the dates, the creator, the title, the description, any tags or taxonomy, any fields for workflow. And this often manifests as content types or a metadata schema. In this screenshot I have, this is actually a content model with different content types. I have the metadata that's needed, the field descriptions, the type of field in use. And the homepage is a content type, and individual event is a content type. So we can see that each of these content types has a title, a description, an author. It has some categories. And you can see which content types these fields apply to. We have special fields for events that only apply to the events content type. Theresa Putkey What we can do with things when they're in fields is we can lay out information using a style sheet instead of manually adjusting the layout on the page. So on the left hand side, we have all of these fields indifferent location in the content management system/ They're not just text on a page. They are separated out. So if we wanted to change the layout of this page, we would do that with the style sheet, instead of manually changing each page by hand. This screenshot shows us the metadata on a services page. So we have page metadata, the title of the description, we have taxonomy driven fields for health topics for services. We can see what type of field that is, if it's required, how many times we can use it. So again, these are just fields on pages and they can be different kinds of fields. Theresa Putkey A subset of those fields can also be taxonomy driven. So taxonomy is a controlled list of terms used to categorize or tag content and assets. If you want to know more about this, you can look up controlled vocabulary, taxonomy, or ontology. And if you're interested in AI, you can look up knowledge graphs, which are very similar to ontologies. So in the previous screenshot, we saw that some of the fields were taxonomy fields, and all we're saying there is that those fields you have to pick from a list of terms to enter something into that field. In an organization, you can have different taxonomies in this organization. They had relevant categories or taxonomies. They had topics, content types. They had a list of industries and the topics that they talked about. They talked about bio products, health and safety, management and operations. And these are controlled lists that they want to use on their site. Theresa Putkey For this project they had, they wanted to do things with projects and topics and events. So they had numerous projects on their site. They had content pages, they had events, they had news for all of these projects. And they wanted to keep track of the projects as a controlled list so they could automatically display content on a page. So say that you have a projects landing page for this ed tech projects, you might have events and news related to that ed tech product projects. And all of that content can be automatically displayed on that tech projects landing page when it's tagged with this project name. Here we have a health topics taxonomy. We have, this is just one level taxonomy. We have in column A, we have the preferred terms that we want to use, we have any synonyms, so in the flu is also called influenza. This taxonomy is targeted at non medical professionals. So we opt for the flu instead of influenza, because the flu is a more colloquial name. We can also say that anytime we're talking about heart disease and heart, I think it's protection. We know that we should use the term heart health instead. We can also relate services and locations to health topics. This can help with automatic display of information on our website. So those are our definitions. And now let's switch to the process. Theresa Putkey So I want to contextualize the process of creating a content model, so you know where it fits into the overall project or website, intranet, extranet redesign process. When I work on content strategy and information architecture projects, I normally go through a discovery and strategy and design phase. In discovery, we do user research, stakeholder engagement, and content audits. And in the strategy and design phase, we work on the content strategy, the content model with taxonomy, and metadata, and then the taxonomy development. So the two I'm going to touch on in this talk are the content audit and the content model. But I just wanted to show you where it fits into the process. Theresa Putkey The content audit that we're looking at, there are a number of different ways to do content audits, in that you want to get different things out of them. So if you're looking at editorial content audit, you would look at different things than a structural content audit, or say, a governance content audit. In this talk, we're going to talk about content audits that create a structural view of the current content. So we look at page structure and content types, available metadata and taxonomies. We look at keywords in the content, on site search terms, by which I mean, when users come to a website, and use the search box that's on the website, those are the on site search terms. Search engines search terms that bring the user to the site. So if somebody is on Google or Bing, they type in these words into the Google search box, and it brings users to your site. We also look at SEO targeted keywords and analytics. Theresa Putkey So here's an example of a content audit, partial content audit. Here we have the URL, the different levels that are on the site, and and their titles. We also look at the different kinds of content types. So this gives us a current view of the different layouts and different types of pages on the site. We look at the keywords being used in the content, and this helps build out the taxonomy later. We can look at the analytics to see the pageviews. Are people visiting this page? How frequently? How much time do they spend on the page? And are there any other issues with it? So you can use a content audit to look at many things like readability, accessibility, but for this structural content audit, we look at structural things. Theresa Putkey We also look at the taxonomies that are available on the site. On the left hand side you'll see that this site uses both categories and tags. This happens to be from a WordPress site. And you can see that the counts for these tags aren't very high. And we have duplicates like accessibility and accessible. Adapting open textbook and adaptation. So we can see that there's really no control over the words that are being typed in here, and therefore we can't do anything with them. On the right hand side is an example of on site search terms. So when somebody comes to the site and types in a term, this is what appears. And it just so happens that on this site, the top search term is physical geology, and project management. And this happens to be an education site. So we use this to help inform the taxonomy development later. Theresa Putkey In our content, audit findings, say we've been through this whole thing and we have found that there's poor structure and navigation on the site. There's inconsistent page layouts, most likely because they're controlled by hand. There's no stylesheet that dictates the layout. There's no fields that people are using in the back end of the content management system. They're just creating a generic page and laying it out however they want. We also might find that there's an inconsistent use of tags or taxonomy, and metadata. So the (...) are that all of these audit findings and the user research and the stakeholder interviews that we've done, have done, point to a couple problems. One, it is hard to find content, both for users and authors. And it might be that there's not enough information or metadata on the content to enable the user experience. So what do we do? How do we create a content model? Theresa Putkey Revising the content model, what I normally do is I ask, given all the problems and organizational goals, what's the best solution for the content model? We always need something more flexible. But we also need more metadata and taxonomy. The more we know about content, the more we can do with it. And I always, always love to start with identifying the current content model and the relationships between the content objects in the content model. So the example I have is for a healthcare company, and we did our content audit, our user research, our stakeholder interviews, and we came up with a view of how the website looked at content, or presented content. Theresa Putkey We found that on the website and really in the organization, that the departments were front and center. You know, people got services from departments. And they had to interact with departments. And a lot of the content on the website was about departments. They had information on services and locations. And they also had information on health topics, but there wasn't any link between the health topics and services and locations. They also hosted some events. And these most likely were not related to anything else on the site. So what are the problems with this model? Well, the problem is that they need to know so much more about their content before they can do anything with it. And what do they need to know? Well, let's look at a revised (...) Theresa Putkey So here, at the very least, they need to know about relationships between the content model, the content objects in the content model. So each of these items is a content object, which will most likely become a content type in your content management system. And the whole thing together makes up a content model. We can see that the departments offer services, and departments are physically located in specific locations. Services are also offered in locations and they're related to health topics. There are events that are about health topics, and events happen in specific locations. So at the very least, we want to know more about the relationship between these content objects. But what else do we want to know? Theresa Putkey Well, we want to know things about those content objects themselves. So here we've added on a layer. So let's see health topics are about one or more topics. services are one of a type of service. So we might have laboratory services or the oratory services. And there can be multiple pages about laboratory services. We might have X ray and imaging and there could be multiple services for X ray and imaging. We have departments and we have different types of department. We have departments are physically located in locations and locations are a type of location. So it's an imaging location. It's a laboratory location is a hospital. So we have multiple hospitals. And we want to know that they're all the same thing, a hospital, we have multiple imaging locations, and we want to know that they're all imaging locations. Theresa Putkey We could also have different types of events as well. But what else is the issue here? Theresa Putkey Earlier I mentioned that this is a department centric model. But we work in user experience and we want to create a user centered model. How do we do that? And where does that lead us? So if we look at the citizen in Canada, I live in Canada. Healthcare is a public institution and so we don't have customers. We might have clients but I'm gonna call them citizens for this discussion. If we want to take a citizen or user centered approach, then what does that mean? What are citizens interested in? Well, citizens are interested in seeing a physician. They see multiple physicians, they're interested in multiple healthcare topics. They might have zero to many diseases and conditions. They want to attend services and that are provided by departments. They might visit many health care locations. They live in a specific community and they can attend different events. So this is quite different, that the departments are kind of tertiary to the discussion. Citizens are most interested in the services they need, the physicians they see and perhaps the diseases and conditions they have they visit many locations and hey, maybe they want a map of that location. Theresa Putkey And there are also other relationships too. And that is kind of layering in that first one where we identified that department centric relationship that departments provide services in many locations. communities, our locations are located in specific communities, and events take place in communities. We see that physicians are experts in different health care topics, and that might be really interesting to us. So it's not just the citizen centric view. But once we illustrate the citizen centric view, then we can also illustrate the different connections between other content objects So where does this lead us? Well, we are talking about what taxonomy and metadata Do we need in our content model. So if this current view is our content model, the whole thing is our content model. The different colored sticky notes are the content objects, which become content types in our content management system.And we have these relationships that illustrate the connection that we need to make in the content management system. Theresa Putkey So what does that leave us we have the health topic content object, the service content, object, location, content, object and community content object. In the health topic, we want to know about the title, the description, the dates, and we have topics here that is a taxonomy field. In the services, we have similar stuff, plus we have the service type. We have it offered in specific locations and we have related topics. So say a service type is, I'm sorry, a service is about heart health, or it's related to heart health, and heart health is one of our health topics. If we have a heart health landing page, then we can automatically show the services related to heart health through the use of taxonomy. Because we've controlled that list. We're applying it consistently. We know that this service is about the heart health topic, then we can show that service on the heart health topic landing page. The location metadata, we'd have location type services offered at that location, related topics. Locations might are Most likely right now, they are in a specific local physical location. And that specific location is within a community. People are when we did the user research, we found out that people were interested in knowing more about their community about knowing the locations, the services and the event in their community. So, community became a content object. And it had geographical boundaries, GPS coordinates and related events. Among other things, it could also have locations or services. Theresa Putkey Let me do this slide is that I modeled what the current framework or current content model was, and I said what our relationships are missing, and what controlled lists are missing. based on you know, what the users need and what the stakeholders need. And then I said, What's the problem with this model overall. And the problem was that it wasn't user centered. So we came and we created a citizen or user centered content model that mapped relationships. And it highlighted the different controlled lists that I need on those different content objects. Then I'm able to list out the metadata that I want to know for each of these content objects. And I can see which fields which metadata fields should be controlled lists. And I know that from identifying these types of services or health topics, that these are also going to be the controlled lists that I need. So just to reiterate, you can figure out your control lists from doing, you're figuring out what metadata you need, and seeing which ones of those are, should be controlled. Plus, you can use this content model, identify your content types and figure out what you want to know about these content types and control the list. Identify your list to control here. And these controlled lists become taxonomy explains some of how to use content modeling to drive your taxonomy framework. Once you figure out your taxonomy that you need, then you can start building it out based on the content audit keywords that you did any SEO implications that you need to consider. And any on site search terms or external search engines search terms. Theresa Putkey Good luck and happy content modeling. Transcribed by https://otter.ai